Kinetics of spontaneous residual shunt closure: a comparative analysis between two surgical procedures for perimembranous ventricular septal defect.

Background: Residual shunt (RS) is common after surgical closure of perimembranous ventricular septal defect (VSD), with smaller size predicting higher spontaneous closure rates. However, most studies focus on final closure rather than the temporal dynamics of closure, and it remains unknown whether conventional surgery and minimally invasive closure differ in RS closure kinetics. This study aimed to investigate the kinetics of spontaneous RS closure by comparing the timing and temporal dynamics between the two surgical approaches. Methods: A retrospective cohort study was conducted. Patients with an RS were enrolled between January 2015 and January 2022, and required to undergo regular outpatient follow-up for 2 years. The patients received surgery with either cardiopulmonary bypass (CPB) or transthoracic minimally invasive closure (MIC) under the guidance of transesophageal echocardiography (TEE). Patients' characteristics, along with pre- and postoperative echocardiography data, were documented. To analyze closure kinetics, Kaplan-Meier survival curves were used; receiver operating characteristic (ROC) analysis was performed for each group to calculate cutoff RS size. This analysis was further stratified by the RS size cutoff values. Additionally, the instantaneous rate of closure over time was assessed using density estimation plots, and the cumulative rate of closure was plotted to visualize cumulative spontaneous closure over time. Results: A total of 168 patients were included; the time to spontaneous closure was significantly shorter in the CPB group [2 months; interquartile range (IQR), 1-4.5 months] than in the MIC group (4 months; IQR, 1-7 months) (P=0.01). RS size emerged as the primary determinant independent predictor in both groups, with a cutoff value of 2.0 mm in the CPB group and 2.1 mm in the MIC group. Kinetic analysis revealed distinct patterns: Overall long-term spontaneous closure rates were similar (log-rank P=0.39). Closure events in the CPB group exhibited a sharply peaked density distribution within the first 6 months (87.4% of closures). In contrast, the MIC group showed a broader, more sustained closure pattern within the first 6 months (72.9% of closures). For moderate-sized RS (≥ cutoff value), the CPB group achieved spontaneous closure at a significantly faster rate (log-rank P=0.045). Conclusions: Although RS size is the primary determinant of spontaneous closure, the surgical approach appears to significantly influence the kinetics of RS closure. The CPB group exhibited early, rapid closure despite larger RS, whereas the MIC group showed a more gradual, sustained closure pattern.